Skip to main content

Iris

Always available
Generative AI Product Builder
AI AdvisorArtificial Intelligence & Machine LearningTechnology

Chat with Iris

Ask Iris anything, or start with one of these:

Your first 3 questions are free, no account needed. They're shared across all advisors.

One question per message works best. Up to 4,000 characters.

How it works

  1. 1

    Ask your question

    Type freely, no need to book a slot.

  2. 2

    Get tailored guidance

    Responses are instant, grounded in the AI Advisor’s topic area.

  3. 3

    Keep the conversation going

    Ask follow-ups, or escalate to an expert anytime.

Book an expert instead

About Iris

I focus on the specific craft of shipping LLM powered features: prompting, retrieval, and agent design that hold up past a demo. I will ask to see your actual failure cases before offering an opinion, because most GenAI problems are eval problems in disguise. Best for people already building, not exploring. Best fit if you are already shipping and hitting a specific wall, not exploring what is possible.

  • Prompt engineering patterns
  • RAG architecture
  • Agent and tool use design
  • Evaluation of LLM outputs
  • Choosing between models and providers
  • Structuring evaluation datasets
  • Reducing latency and cost per request
  • Choosing chunking and retrieval strategy for RAG
  • Designing guardrails for user facing AI features
Iris is an AI system, not a real person. Responses are for general guidance.